This is the largest one I know of: http://icofy-blog.de/en/icofy-base/
The Millionbase 2.2 is a little smaller (PGN format): http://www.top-5000.nl/pgn.htm
Millbase is another slightly smaller database: http://katar.weebly.com/2/post/2013/08/millbase-for-scid-update3.html
Does biggest necessarily = best?
Well the top link says that with each update, it works on removing duplicates of the same game. Also, quality v quantity?
Does biggest necessarily = best?
Uh, duh, yes.
NO!!! If you want a quality database check out the 2013 openings by chessbase or there megadatabase. These are analyzed quality games.
Not saying the ones suggested are poor. Yet there are some crummy databases out there.
Does biggest necessarily = best?
Possibly the opposite. I use ChessBase online and Big 2012 every day (the difference between Big and Mega is not the games, but annotations to the games). When searching a position, I find far too many games played by weak players.
A database that has only master games would be much smaller, and much better quality. Of course, with the larger database, I can create the smaller thanks to the features of the ChessBase database program.
The ONLY advantage a really big DB has, is that it plays out opening traps.
If you're researching eg the 2 Knights Fritz Variation, a DB of Super GM games is not gonna tell you what should happen after 6 d6. A patzer DB will.
And the major disadvantage of a large database is that the percentages can be terribly misleading. Consider the Fried Liver Attack: it is not played by masters because Black is equal or slightly better with correct defense. Nonetheless, White scores 70% wins in the database.
And the major disadvantage of a large database is that the percentages can be terribly misleading. Consider the Fried Liver Attack: it is not played by masters because Black is equal or slightly better with correct defense. Nonetheless, White scores 70% wins in the database.
Oh, I agree asolutely.
So are there any good free options or should I save up for Chessbase? I prefer to have an application vs. a website database.
When playing correspondence Chess, it is a good idea to focus on games played in the last 5 years in one`s database. Refine your search to players rated over 2400 (both of them). Play the percentages and study most of those selected games. Some of those games may be marred by silly blunders toward the end of time controls so the various lines that look good yet lead to losses are still really quite good. Then, start looking at the games of the lesser players (still using the 5 year criteria and the other methodology that I cited). Then, if you have time, go back in time and do the same in 5 year increments and even look at even games with 1800 strength players. That should be an intelligent usage of databases. Granted, this is time consuming but it really pays off. I know because I have tied for second in most of the events and have at least 80 games under my belt. I`ve won a double round robin, been into not only in a semi finals of the USCF`s Electronic Golden Knights but advanced into a finals. True, I only scored 1.5 out of 6 in that, but then again, I was seeded no.7 and all games went down to the wire. I also have a 2000+ postal rating. I use ChessBase 12 and after my games are over, win, lose or draw, I analyse with Houdini Pro 3. That 5 year guideline is based on the fact that it seems that major evaluations of openings/defenses tend to follow this timeline for the most part.
Excellent post. Thanks. Your methods differ quite a bit from those I use. Are they more effective? That's the discussion that I keep trying to provoke when I start threads on databases.
http://www.chess.com/forum/view/general/using-databases-well
http://www.chess.com/blog/Ziryab/playing-with-databases
When playing correspondence Chess, it is a good idea to focus on games played in the last 5 years in one`s database. Refine your search to players rated over 2400 (both of them). Play the percentages and study most of those selected games. Some of those games may be marred by silly blunders toward the end of time controls so the various lines that look good yet lead to losses are still really quite good. Then, start looking at the games of the lesser players (still using the 5 year criteria and the other methodology that I cited). Then, if you have time, go back in time and do the same in 5 year increments and even look at even games with 1800 strength players. That should be an intelligent usage of databases. Granted, this is time consuming but it really pays off. I know because I have tied for second in most of the events and have at least 80 games under my belt. I`ve won a double round robin, been into not only in a semi finals of the USCF`s Electronic Golden Knights but advanced into a finals. True, I only scored 1.5 out of 6 in that, but then again, I was seeded no.7 and all games went down to the wire. I also have a 2000+ postal rating. I use ChessBase 12 and after my games are over, win, lose or draw, I analyse with Houdini Pro 3. That 5 year guideline is based on the fact that it seems that major evaluations of openings/defenses tend to follow this timeline for the most part.
Excellent post. Thanks. Your methods differ quite a bit from those I use. Are they more effective? That's the discussion that I keep trying to provoke when I start threads on databases.
http://www.chess.com/forum/view/general/using-databases-well
Thanks for the compliment; I`ll check out your posts about database usage. Do you play in the USCF events,OTB or Correspondence? You could reply on this site.
I play in several USCF OTB events every year in my city, and every few years out of town. I played USCF correspondence 1997-2000, and 1978-1980.
I did find a copy of TWIC's archive and I am going to use that as my database for now.
I've got a question. I found out the archive I have is TWIC 1-957. It is also read-only. I've downloaded TWIC 958-current in pgn. How would I go about creating a database that included all of them?
So, does anyone know how to combine games together to create a database in scid or scid v pc?
I answered my own question, just wanted to add an update in case this would help anyone.
Here is a youtube video: http://www.youtube.com/watch?v=vwdkJg9jrIo
The file I downloaded with the TWIC archive was already in SCID format. So I just needed to open that and then go under tools, import pgn, and then select all the rest of the issues that I had downloaded.
Boom! 1,527,317 games. :)
I just downloaded SCID v PC. Where can I find the biggest database of games to download for it?